AI Info on your posts will now be penalized by the algorithm on Meta (Facebook and Instagram), and other platforms are following.
I hate to say I told you so… But, I am a passionate person when it comes to content and original thought because for many of my clients, their voice is their currency. Most of my authors and thought-leaders have been ahead of the curve on AI, viewing it as an idea-generating tool to think through concepts before writing. It does not write for them.
In the words of my long-time mentor, “Writers, write.” (miss you, DG)
And for my non-writers… do not get mission-driven or creative brands started on using AI instead of doing photography, videography, and writing. I’ve had countless conversations around this with clients, and man, am I proud to call these thinkers mine.
Don’t get me wrong, I experiment with AI and see how or where it can be used for myself and clients, but some red flags have been there for a while, and we need to protect our brands.
Here is what’s happening:
- Facebook and Instagram (Meta) are adding AI INFO to your posts, and when people click on it, it tells users that your content is completely AI-generated.
- On YouTube, it takes center stage in your description.
- You have an option to disclose manually on Meta, but your content will be restricted either way.
- And if you don’t disclose, they will catch it and label it.
- If you use AI too often, even your original content will be restricted.
- More platforms are following.
What this means for you:
- If you have completely AI-generated content, your credibility is going to tank.
- The algorithm will literally restrict your content as they continue to crack down.
- Your Ad Manager account could get shut down or restricted.
- Your business or creator pages could be deleted or restricted.
- Meta has a history of making implementations that they have to refine over time, so you’re in the danger zone if you’re using too much AI.
How to not use AI:
- Do not mass generate social content or emails with AI – yes, even your emails can start getting flagged as spam.
- Pump the brakes on social posts entirely written by AI. Yes, even if you did the work on the prompt.
- Reconsider AI images and videos.
How to use AI:
- The better the prompt, the better the output with AI. So use it as an idea generator and refiner, then do the work to write and create.
- Be cautious with AI-edited images/video or completely AI-generated images/video. This is a fun feature, and doesn’t have to be used maliciously, but it’s a good idea to scale way back because there are new penalties.
Remember, when it comes to AI, it has the pesky ability to “hallucinate.” If you haven’t heard that term, it’s a good time to check it out. When you ask AI to pull quotes from a transcript, provide data, or perform other tasks that require hard info, it can fabricate information and present it as fact. I’ve called out ChatGPT many times, testing it by uploading transcripts, studies, and other information that I’d love to use for collateral quickly. But I learned FAST that it gets it wrong… a lot.
This is all excellent news for any mission-driven brand. Why? Because typically, building authority, being genuine, and showing up in a way that feels good is vital to you. It’s going back to the fundamentals – we don’t need to mass produce; we need to focus on quality.
Quality and originality always win in the long run.
Want to know more about producing content that isn’t gimmicky? Check out my new little book…
From Meta:
When content will have a label
- When Meta AI tools are used. Content that they’ll edit or modify using Meta’s AI tools and share on your Facebook directly as a post, story, or reel may automatically be labeled above or on the content. For example, some content may have a visible watermark that says Imagined with AI.
- When created content has AI signals. Content with industry-standard signals indicating that it was created with AI info. This includes content that is created using third-party AI tools. Content that is created using Meta’s AI tools, downloaded to a device and then uploaded to Facebook will also receive an AI info label.
Content with industry-standard signals indicating that it was modified with AI will not receive an AI info label directly on the post, story, reel or thread.
Information about AI used to create or modify content can be found by clicking Options on any post, story, reel or thread, regardless of the type of signal Meta receives.
This approach will evolve as people’s expectations and the technology evolve.
When the AI info label is required
Meta requires you to label content you share that has photorealistic video or realistic-sounding audio that has been digitally generated or altered, including with AI. This means that if a digital creation or AI tool was used to create or modify this kind of content, you must label the content before you share it.
Meta does not require you to label images that have been created or modified with AI. Images will still receive a label if Meta’s systems detect they were AI-generated.
Note: There may be penalties if you do not label content as required.
Here are some examples of digitally created content that requires a label:
- A video that appears realistic of a group of people walking around an outdoor market
- An audio file of two people talking
- A song created using AI-generated vocals
- A reel narrated with a realistic AI-generated voiceover
Here are some examples of digitally created content that does not require a label:
- Video of an outdoor landscape, created in a style resembling a cartoon
- Image of person riding a bull (Meta does not require you to label images)
